The DOI matches its Crossref record; 61 of the 61 references it cites were found in a registry or an index.
Separating Style and Content with Bilinear Models
Abstract
Perceptual systems routinely separate “content” from “style,” classifying familiar words spoken in an unfamiliar accent, identifying a font or handwriting style across letters, or recognizing a familiar face or object seen under unfamiliar viewing conditions. Yet a general and tractable computational model of this ability to untangle the underlying factors of perceptual observations remains elusive (Hofstadter, 1985). Existing factor models (Mardia, Kent, & Bibby, 1979; Hinton & Zemel, 1994; Ghahramani, 1995; Bell & Sejnowski, 1995; Hinton, Dayan, Frey, & Neal, 1995; Dayan, Hinton, Neal, & Zemel, 1995; Hinton & Ghahramani, 1997) are either insufficiently rich to capture the complex interactions of perceptually meaningful factors such as phoneme and speaker accent or letter and font, or do not allow efficient learning algorithms. We present a general framework for learning to solve two-factor tasks using bilinear models, which provide sufficiently expressive representations of factor interactions but can nonetheless be fit to data using efficient algorithms based on the singular value decomposition and expectation-maximization. We report promising results on three different tasks in three different perceptual domains: spoken vowel classification with a benchmark multi-speaker database, extrapolation of fonts to unseen letters, and translation of faces to novel illuminants.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“GeAN also uses the bilinear model [29,30] to distil the knowledge interwoven in these representations by allowing each node and edge to collaboratively modulate each other's contributions through a shared learning space.”
From TVGeAN: Tensor Visibility Graph-Enhanced Attention Network for Versatile Multivariant Time Series Learning Tasks (2024) · cites this paper as [29]
Citations
Cites
Maximum Likelihood from Incomplete Data Via the EM Algorithm
Dempster, A. P., Laird, N. M., & Rubin, D. B. (1977) · Journal of the Royal Statistical Society Series B: Statistical Methodology · doi:10.1111/j.2517-6161.1977.tb01600.x · Cited by 50,444
Neural networks for pattern recognition
Bishop, C. (1994) · Choice Reviews Online · doi:10.5860/choice.31-5500 · Cited by 18,537
Press, W. H., Teukolsky, S. A., Vetterling, W. T., & Flannery, B. P. (1994) · Cited by 15,411
Turk, M., & Pentland, A. (1991) · Journal of Cognitive Neuroscience · doi:10.1162/jocn.1991.3.1.71 · Cited by 13,733
(1973) · CERN Document Server (European Organization for Nuclear Research) · Cited by 12,895
Numerical recipes in Pascal: the art of scientific computing
(1990) · Choice Reviews Online · doi:10.5860/choice.27-5158 · Cited by 12,040
An Information-Maximization Approach to Blind Separation and Blind Deconvolution
Bell, A. J., & Sejnowski, T. J. (1995) · Neural Computation · doi:10.1162/neco.1995.7.6.1129 · Cited by 8,986
Shape and motion from image streams under orthography: a factorization method
Tomasi, C., & Takeo Kanade, A. (1992) · International Journal of Computer Vision · doi:10.1007/bf00129684 · Cited by 2,879
Matrix Differential Calculus with Applications in Statistics and Econometrics.
Magnus, J. R., & Neudecker, H. (1988) · Biometrics · doi:10.2307/2531754 · Cited by 2,729
Classical Electrodynamics (2nd edn)
Brereton, M. (1976) · Physics Bulletin · doi:10.1088/0031-9112/27/8/041 · Cited by 2,612
Cited by
Sorted by citations
Items 401–500 of 839
A style controller for generating virtual human behaviors
Chiu, C., & Marsella, S. (2011) · doi:10.5555/2034396.2034415 · Cited by 14
A Bilinear Model for Sparse Coding
Grimes, D. B., & Rao, R. P. N. (2002) · Cited by 14
Tensor Component Analysis for Interpreting the Latent Space of GANs
Oldfield, J. P., Georgopoulos, M., Panagakis, Y., et al. (2021) · British Machine Vision Conference (BMVC) · doi:10.5244/c.35.430 · Cited by 13
Modeling the Lighting in Scenes as Style for Auto White-Balance Correction
Kınlı, F., Yilmaz, D., Özcan, B., et al. (2023) · IEEE Winter Conference on Applications of Computer Vision · doi:10.1109/wacv56688.2023.00488 · Cited by 13
“With the help of this idea, the content of an observation can be expressed in a more compact way [63, 44, 24, 43, 63], the style of an observation can be recognized [23, 48, 66] and manipulated to the desired style [32, 36, 46, 53] or even novel contents can be generated with a particular style [33, 41, 51, 49, 50].”
Li, X., & Huan, J. (2017) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3097983.3097994 · Cited by 13
Bilinear CNN Model for Fine-Grained Classification Based on Subcategory-Similarity Measurement
Dai, X., Gong, S., Zhong, S., & Bao, Z. (2019) · Applied Sciences · doi:10.3390/app9020301 · Cited by 16
“Tanenbaum and Freeman [27] proposed a bilinear model to model the change of two-factor variations, such as the “style” and “content” of the image.”
BackgroundGu, S., Nojima, Y., Ishibuchi, H., & Wang, S. (2019) · IEEE Transactions on Fuzzy Systems · doi:10.1109/tfuzz.2019.2906855 · Cited by 12
“As a whole, the implicit style correlations between a pair of data samples, in addition to physical features of data, may be greatly useful for data classification [4]–[6].”
BackgroundLiang, M., Du, J., Yang, C., et al. (2019) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2019.2945567 · Cited by 13
“…approaching the cross-media search problem include canonical correlation analysis (CCA) [4], partial least squares (PLS) [5], bilinear model (BLM) [6], CCQ [7], collective deep quantization (CDQ) [8], collective matrix factorization hashing (CMFH) [9], and coupled feature space-based learning…”
BackgroundZhao, B., Tang, W., Pan, Y. R., Han, Y., & Wang, W. (2021) · Electronics · doi:10.3390/electronics10172046 · Cited by 14
“Considering the associated inheritance of advantages of both concatenation and element-wise multiplication [54], we implement bilinear pooling after the discriminant MRELM-AE to extract higher-order features, thereby enhancing the discrimination of the features.”
MethodPoly-NL: Linear Complexity Non-local Layers With 3rd Order Polynomials
Babiloni, F., Marras, I., Kokkinos, F., Deng, J., Chrysos, G. G., & Zafeiriou, S. P. (2021) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.01035 · Cited by 16
“In Bilinear layers [46, 6, 32, 55] feature maps of different networks get bilinearly combined together to capture pairwise interactions.”
BackgroundResidual Tensor Train: A Quantum-Inspired Approach for Learning Multiple Multilinear Correlations
Chen, Y., Pan, Y., & Dong, D. (2022) · IEEE Transactions on Artificial Intelligence · doi:10.1109/tai.2022.3194132 · Cited by 17
Person identification from action styles
Kviatkovsky, I., Shimshoni, I., & Rivlin, E. (2015) · IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw.2015.7301323 · Cited by 13
“Although reminiscent of the content-style separation paradigm [19], our generative model does not assume any specific dependence, e.”
BackgroundMethodDeformation-Based 3D Facial Expression Representation
Demisse, G., Aouada, D., & Ottersten, B. (2018) · ACM Transactions on Multimedia Computing Communications and Applications · doi:10.1145/3176649 · Cited by 13
“In Mpiperis et al. (2008) and Tenenbaum and Freeman (2000), observed facial data are formulated as a bilinear function of face and expression.”
BackgroundAn Animal Classification based on Light Convolutional Network Neural Network
Jiang, B., Huang, W., Tu, W., & Yang, C. (2019) · International Conference on Intelligent Computing and Its Emerging Applications (ICEA) · doi:10.1109/icea.2019.8858309 · Cited by 14
Informative Multimodal Unsupervised Image-to-Image Translation
Doan, T. T., Ghyselinck, G., & Hanczar, B. (2021) · doi:10.5121/csit.2021.110503 · Cited by 13
“In this work, we disentangle this style code by maximizing the mutual information between this code and the generated image.”
MethodThe Statistical Inefficiency of Sparse Coding for Images (or, One Gabor to Rule them All)Preprint
Bergstra, J., Courville, A. C., & Bengio, Y. (2011) · arXiv (Cornell University) · doi:10.48550/arxiv.1109.6638 · Cited by 13
“Bilinear modeling [Tenenbaum and Freeman, 2000, Grimes and Rao, 2005], z ∼ N m∑ i=1 n∑ j=1 wijxiyj , σ 2 , (1) has been put forward as a model for separating content from style.”
MethodLearning Typographic StylePreprint
Baluja, S. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1603.04000 · Cited by 13
Incremental multi-linear discriminant analysis using canonical correlations for action recognition
Jia, C., Wang, S., Peng, X., et al. (2011) · Neurocomputing · doi:10.1016/j.neucom.2011.11.006 · Cited by 12
“Most traditional algorithms, such as principal component analysis (PCA) [6] [7] and linear discriminant analysis (LDA) [8] [9], represent an object as a 1-dimension vector.”
MethodButail, S., Bollt, E. M., & Porfiri, M. (2013) · Journal of Theoretical Biology · doi:10.1016/j.jtbi.2013.07.029 · Cited by 12
Huang, K., Jiang, H., & Zhang, X.-Y. (2017) · doi:10.1145/3109761.3158392 · Cited by 12
“Moreover, related work investigated in [1], [2] are computationally inefficient.”
BackgroundMethodResultInfluentialCross-Modality Personalization for Retrieval
Murrugarra-Llerena, N., & Kovashka, A. (2019) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr.2019.00659 · Cited by 12
“In prior work, [52] separate content and style for handwritten Chinese characters, by training separate networks for each, and [38] use a linear model in both the content (character ID) and handwriting style.”
BackgroundMulti-View Wasserstein Discriminant Analysis with Entropic Regularized Wasserstein Distance
Kasai, H. (2020) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp40776.2020.9054427 · Cited by 11
“They include, for example, partial least squares (PLS) [6], canonical correlation analysis (CCA) [7, 8], bilinear model (BLM) [9], multi-view discriminant analysis (MvDA) [10], generalized multi-view analysis (GMA) [11], and standard linear multi-view discriminant analysis (S-LMvDa) [12].”
MethodLearning Visual Styles from Audio-Visual Associations
Li, T., Liu, Y., Owens, A., & Zhao, H. (2022) · Lecture notes in computer science · doi:10.1007/978-3-031-19836-6_14 · Cited by 12
Homeomorphic Manifold Analysis: Learning Decomposable Generative Models for Human Motion Analysis
Lee, C.-S., & Elgammal, A. (2007) · Springer eBooks · doi:10.1007/978-3-540-70932-9_8 · Cited by 12
“Multilinear tensor analysis decomposes multiple orthogonal factors as an extension of principal component analysis (PCA) (one orthogonal factor), and bilinear model (two orthogonal factors) [ 25 ].”
MethodLow-Rank Bilinear Classification: Efficient Convex Optimization and Extensions
Kobayashi, T. (2014) · International Journal of Computer Vision · doi:10.1007/s11263-014-0709-5 · Cited by 12
Kernelizing Sorting, Permutation, and Alignment for Minimum Volume PCA
Jebara, T. (2004) · Lecture notes in computer science · doi:10.1007/978-3-540-27819-1_42 · Cited by 12
“These tools have been used for finding correspondence and aligning images of, for instance, digits [2, 14 ] to obtain better models (such as morphable or corresponded models).”
MethodAnalysis, Synthesis, and Retargeting of Facial Expressions
Bregler, C., & Chuang, E. S. (2004) · Cited by 12
Subspace learning by kernel dependence maximization for cross-modal retrieval
Xu, M., Zhu, Z., Zhao, Y., et al. (2018) · Neurocomputing · doi:10.1016/j.neucom.2018.04.073 · Cited by 12
Two-stage deep learning for supervised cross-modal retrieval
Shao, J., Zhao, Z., & Su, F. (2018) · Multimedia Tools and Applications · doi:10.1007/s11042-018-7068-0 · Cited by 12
“Other algorithms are also proposed to deal with cross-modal retrieval problems, such as partial least square (PLS) [24], Bilinear Model(BLM) [18, 21, 38] and etc.”
BackgroundProbabilistic bilinear models for appearance-based vision
Grimes, Shon, & Rao, R. P. N. (2003) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2003.1238665 · Cited by 12
“Previous results using the bilinear model [14] used a sum squared error (SSE) criterion in developing learning and inference procedures.”
BackgroundMethodResultInfluentialMulti-View Face Image Synthesis Using Factorization Model
Du, Y., & Lin, X. (2004) · Lecture notes in computer science · doi:10.1007/978-3-540-24837-8_19 · Cited by 12
Gooßen, A., Hermann, E., Weber, G. M., et al. (2010) · Computer Science - Research and Development · doi:10.1007/s00450-010-0139-8 · Cited by 11
“In the recent years bilinear modeling [ 34 ] has risen in popularity.”
Linear Complexity Self-Attention with $3^{\text{rd}}$ Order Polynomials
Babiloni, F., Marras, I., Deng, J., Kokkinos, F., Maggioni, M., Chrysos, G. G., Torr, P. H. S., & Zafeiriou, S. P. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2022.3231971 · Cited by 11
“In Bi-linear layers [21], [22], [23], [24] feature maps of different networks get bilinearly combined together to capture pairwise interactions.”
BackgroundA novel illumination-robust face recognition using statistical and non-statistical method
Jun, B., Lee, J., & Kim, D. (2010) · Pattern Recognition Letters · doi:10.1016/j.patrec.2010.09.011 · Cited by 11
Motion Models for People Tracking
Fleet, D. J. (2011) · Springer eBooks · doi:10.1007/978-0-85729-997-0_10 · Cited by 11
“, [53, 65, 72]), one might consider a weighted sum of GPs, {gi(x)}:”
BackgroundLMDT: A weakly-supervised large-margin-domain-transfer for handwritten digit recognition
Hosseinzadeh, H., & Razzazi, F. (2016) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2016.02.014 · Cited by 11
Deep hash for latent image retrieval
Fanfeng, Z., Hu, S., & Xiao, K. (2019) · Multimedia Tools and Applications · doi:10.1007/s11042-019-07980-9 · Cited by 11
Tensor Decomposition and Non-linear Manifold Modeling for 3D Head Pose Estimation
Derkach, D., Ruiz, A., & Sukno, F. M. (2019) · International Journal of Computer Vision · doi:10.1007/s11263-019-01208-x · Cited by 11
“…by feature vector x) whose rotation is ω(y)0 , we can obtain its identity vector 2 This process is coined translation in the seminal work by Tenenbaum and Freeman (2000). u(id) and combine it with f(y) to synthesize unknown views of the same object according to different rotation angles…”
Face Recognition from Video by Matching Image Sets
Chin, T.-J., James, U., Schindler, K., & Suter, D. (2005) · International Conference on Digital Image Computing: Techniques and Applications · doi:10.1109/dicta.2005.36 · Cited by 15
Su, Y., Li, S., Wang, S., et al. (2014) · IEEE Transactions on Circuits and Systems for Video Technology · doi:10.1109/tcsvt.2014.2329375 · Cited by 11
“One special case is the extended ISOMap [46], which makes use of a new feature, representing each data point by a feature vector of its geodesic distance to any points, to make sure new samples can be evaluated by the training samples.”
BackgroundInfluentialSufficient Canonical Correlation Analysis
Guo, Y., Ding, X., Liu, C., et al. (2016) · IEEE Transactions on Image Processing · doi:10.1109/tip.2016.2551374 · Cited by 11
Facial Expression Transformations for Expression-Invariant Face Recognition
Lee, H.-S., & Kim, D. (2006) · Lecture notes in computer science · doi:10.1007/11919476_33 · Cited by 11
Generalized Low-Rank Approximations
Srebro, N., & Jaakkola, T. (2003) · DSpace@MIT (Massachusetts Institute of Technology) · Cited by 11
Facial Expression Recognition System
Ren, Y. (2008) · UWSpace (University of Waterloo) · Cited by 11
Nonlinear Generative Models for Dynamic Shape and Dynamic Appearance
Elgammal, A. (2005) · IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvpr.2004.407 · Cited by 11
“I. INTRODUCTION”
Vasilescu, M. A. O. (2009) · TSpace (University of Toronto) · Cited by 11
GlyphGAN: Style-Consistent Font Generation Based on Generative Adversarial NetworksPreprint
Hayashi, H., Abe, K., & Uchida, S. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1905.12502 · Cited by 11
“One of the classical methods is example-based font generation [8, 9, 10, 11, 12, 13].”
MethodRelatively-Paired Space Analysis: Learning a Latent Common Space From Relatively-Paired Observations
Kuang, Z., & Wong, K.-Y. K. (2014) · International Journal of Computer Vision · doi:10.1007/s11263-014-0783-8 · Cited by 10
LRF-WiVi: A WiFi and Visual Indoor Localization Method Based on Low-Rank Fusion
Liu, W., Qin, C., Deng, Z., & Jiang, H. (2022) · Sensors · doi:10.3390/s22228821 · Cited by 10
“The bilinear pooling model proposed by Tenenbaum and Freeman [28] captures the complex interaction between “content” and “style” through the outer product.”
BackgroundMethodMulti-Modal Machine Learning for Vehicle Rating Predictions Using Image, Text, and Parametric Data
Su, H., Song, B., & Ahmed, F. (2023) · ASME International Computers in Engineering Conference and Engineering Database Symposium · doi:10.1115/detc2023-115076 · Cited by 15
“This approach can capture the high-order multiplicative interactions among all modalities, leading to more expressive and predictive multi-modal representations for fine-grained recognition [42, 44].”
Background11 Movement Style, Movement Features, and the Recognition of Affect from Human Movement
Pollick, F. E., & Paterson, H. (2008) · Oxford University Press eBooks · doi:10.1093/acprof:oso/9780195188370.003.0013 · Cited by 10
Video‐based facial expression recognition by removing the style variations
Mohammadian, A., Aghaeinia, H., & Towhidkhah, F. (2015) · IET Image Processing · doi:10.1049/iet-ipr.2013.0697 · Cited by 10
“We assume f is a asymmetric bilinear model, given as (4) in which the terms w iju vary with style [14] x scu ( t ) = where a = i a si w siju and i , j and u denote the components of style, content and observation vectors, respectively.”
MethodMutual Complementarity: Multi-Modal Enhancement Semantic Learning for Micro-Video Scene Recognition
Guo, J., Nie, X., & Yin, Y. (2020) · IEEE Access · doi:10.1109/access.2020.2973240 · Cited by 10
“The unsupervised methods include Canonical Correlation Analysis (CCA) [14], Partial Least Squares (PLS) [15], Bilinear Model (BLM) [16], and Deep Canonical Correlation Analysis (DCCA) [17], whereas the supervised methods include Generalized Multi-view Analysis (GMA) [18], Multi-view Discriminant Analysis (MvDA) [19],…”
MethodExpression-Compensated 3D Face Recognition with Geodesically Aligned Bilinear Models
Mpiperis, I., Malassiotis, S., & Strintzis, M. G. (2008) · IEEE International Conference on Biometrics: Theory, Applications and Systems (BTAS) · doi:10.1109/btas.2008.4699364 · Cited by 10
“In this paper, we use a symmetric bilinear model [1] to decouple the impact of identity and expression on the appearance of the face surface and encode quantitatively their contribution in independent control parameters.”
BackgroundMethodInfluential2D artistic images analysis, a content-based survey
Hurtut, T. (2010) · Cited by 10
Image normalization by mutual information
Bart, E., & Ullman, S. (2004) · British Machine Vision Conference (BMVC) · doi:10.5244/c.18.35 · Cited by 10
“Similar constraints are present in [21, 23, 20, 17, 2, 10, 25].”
BackgroundImages, Frames, and Connectionist Hierarchies
Dayan, P. (2006) · Neural Computation · doi:10.1162/neco.2006.18.10.2293 · Cited by 10
“A second key antecedent is the work of Tenenbaum and Freeman (2000) on multilinear generative models for the mostly unsupervised separation of different factors (“content,” for us the nature of the face, and “style,” the attentional focus) that interact to determine inputs.”
BackgroundMethodInfluentialEfficiently estimating facial expression and illumination in appearance-based tracking
Buenaposada, J. M., Muñoz, E., & Baumela, L. (2006) · British Machine Vision Conference (BMVC) · doi:10.5244/c.20.7 · Cited by 10
“Subspace approaches have also been used to separate multiple orthogonal factors using bilinear [16, 9] or multi-linear [18] models.”
MethodDisentangled Representations in Neural ModelsPreprint
Whitney, W. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1602.02383 · Cited by 15
Quantitative Evaluation of Style TransferPreprint
Yeh, M.-C., Tang, S., Bhattad, A., & Forsyth, D. A. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1804.00118 · Cited by 11
“Bilinear models are capable of simple image style transfer [6] by factorizing style and content representations, but non-parametric methods like patch-based texture synthesis can deal with much more complex texture fields [7].”
BackgroundA Learning-Based Multi-Node Fusion Positioning Method Using Wearable Inertial Sensors
Song, Y., Xia, S., Yang, J., & Pei, L. (2024) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp48485.2024.10447298 · Cited by 9
“To capture this temporal dependency, the Weight Encoder E W utilizes a two-layer structure of Bidirectional-LSTM [19] with 48 hidden states, allowing for enhanced incorporation of temporal cues derived from IMU data.”
MethodStructured Generative Models for Scene Understanding
Williams, C. K. I. (2024) · International Journal of Computer Vision · doi:10.1007/s11263-024-02316-z · Cited by 9
“Tenenbaum and Freeman (2000) use a bilinear model to separate style and content, e.g. of the font (style) of different letters (content).”
Diverse videos synthesis using manifold‐based parametric motion model for facial understanding
Mohammadian, A., Aghaeinia, H., & Towhidkhah, F. (2016) · IET Image Processing · doi:10.1049/iet-ipr.2014.0905 · Cited by 9
Hu, W., & Hu, H. (2018) · Neural Processing Letters · doi:10.1007/s11063-018-9942-1 · Cited by 9
Improved Bilinear Model for Facial Expression Recognition
Mahmoudi, M. A., Chetouani, A., Boufera, F., et al. (2021) · Communications in computer and information science · doi:10.1007/978-3-030-71804-6_4 · Cited by 9
VE-KWS: Visual Modality Enhanced End-to-End Keyword Spotting
Zhang, A., He, W., Guo, P., Fu, Y., Xie, L., Gao, Y., Zhang, S., & Feng, J. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10096858 · Cited by 9
“After that, we use bilinear pooling, a powerful method for fusing multi-modal representations [23], to fuse the joint representation and uni-modal representation via where W ja and W jv are projecting weight matrices for different modalities.”
MethodChen, S., Gyawali, P., Liu, H., et al. (2017) · IEEE International Symposium on Biomedical Imaging (ISBI) · doi:10.1109/isbi.2017.7950596 · Cited by 9
Facial age feature extraction based on deep sparse representation
Liao, H. (2018) · Multimedia Tools and Applications · doi:10.1007/s11042-018-6342-5 · Cited by 9
Masked Linear Regression for Learning Local Receptive Fields for Facial Expression Synthesis
Khan, N., Akram, A., Mahmood, A., Ashraf, S., & Murtaza, K. (2019) · International Journal of Computer Vision · doi:10.1007/s11263-019-01256-3 · Cited by 9
“We also compare with basis learning approaches including PCA and SVD-based bilinear model for separation of style and content (SSC) (Tenenbaum and Freeman 2000).”
MethodInfluentialFace beautification: Beyond makeup transfer
Liu, X., Wang, R., Peng, H., et al. (2022) · Frontiers in Computer Science · doi:10.3389/fcomp.2022.910233 · Cited by 9
“ied in the literature which dated back to content-style separation [36].”
BackgroundImproving the Fusion of Acoustic and Text Representations in RNN-T
Zhang, C., Li, B., Lu, Z., Sainath, T. N., & Chang, S.-Y. (2022) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp43922.2022.9747418 · Cited by 11
Content-Context Factorized Representations for Automated Speech Recognition
Chan, D., & Ghosh, S. (2022) · Interspeech (USB) · doi:10.21437/interspeech.2022-390 · Cited by 12
Multi-view Face Recognition by Nonlinear Tensor Decomposition
Tian, C., Fan, G., & Gao, X. (2008) · Proceedings - International Conference on Pattern Recognition/Proceedings/International Conference on Pattern Recognition · doi:10.1109/icpr.2008.4761195 · Cited by 9
“Bilinear analysis was introduced in [9] to represent views and identities under a uniform basis and separate “style” (view or pose) and “content” (identity) factors.”
MethodA weakly supervised large margin domain adaptation method for isolated handwritten digit recognition
Hosseinzadeh, H., Razzazi, F., & Kabir, E. (2016) · Journal of Visual Communication and Image Representation · doi:10.1016/j.jvcir.2016.02.018 · Cited by 9
“Tenenbaum and Freeman [16] have employed a bilinear model to separate the style and class knowledge in a group of patterns.”
MethodMultimodal Local Perception Bilinear Pooling for Visual Question Answering
Lao, M., Guo, Y., Wang, H., & Zhang, X. (2018) · IEEE Access · doi:10.1109/access.2018.2873570 · Cited by 10
“Previous VQA frameworks attempted to utilize bilinear pooling [4] methods to establish sufficient interactions between features from two modals for multimodal feature fusion, rather than rely on concatenating vectors or applying element-wise sum or product.”
MethodMulti-class joint subspace learning for cross-modal retrieval
Yu, E., Li, J., Wang, L., et al. (2018) · Pattern Recognition Letters · doi:10.1016/j.patrec.2018.08.012 · Cited by 9
Learning to relate images: Mapping units, complex cells and simultaneous eigenspacesPreprint
Memisevic, R. (2011) · arXiv (Cornell University) · doi:10.48550/arxiv.1110.0107 · Cited by 9
“Like an auto-encoder, this model can be trained with gradient based optimization.”
BackgroundMethodLearning Transposition-Invariant Interval Features from Symbolic Music and AudioPreprint
Lattner, S., Grachten, M., & Widmer, G. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1806.08236 · Cited by 9
“n extension was proposed which involves the correlation of neural patterns. In machine learning, this principle was deployed in bilinear models, for example to separate person and pose in face images [30]. Bi-linear models, like the GAE, are two-factor models whose outputs are linear in either factor when the other is…”
MethodDeep Temporal Linear Encoding NetworksPreprint
Diba, A., Sharma, V., & Gool, L. V. (2016) · arXiv (Cornell University) · doi:10.48550/arxiv.1611.06678 · Cited by 18
“Likewise, Bilinear models [20, 32] have been widely used and have achieved state-of-the-art results.”
MethodMulti‐modal object detection via transformer network
Liu, W., Wang, H., Gao, Q., & Zhu, Z. (2023) · IET Image Processing · doi:10.1049/ipr2.12884 · Cited by 8
Context-Aware end-to-end ASR Using Self-Attentive Embedding and Tensor Fusion
Chang, S.-Y., Zhang, C., Sainath, T. N., Li, B., & Strohman, T. (2023) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49357.2023.10095204 · Cited by 8
“In particular, a low-rank approximation to bilinear pooling with residual connections [24–26] have been used as the joint network [22], which fuses the two representations in a more expressive outer product h acou t ⊗ h pred u .”
MethodFactorization of view-object manifolds for joint object recognition and pose estimation
Zhang, H., El-Gaaly, T., Elgammal, A., et al. (2015) · Computer Vision and Image Understanding · doi:10.1016/j.cviu.2015.03.014 · Cited by 8
“The pioneering work of Tenenbaum and Freeman [36] formulated the separation of style and content using a bilinear model framework.”
MethodA Study of Bilinear Models in Voice Conversion
Popa, V., Nurminen, J., & Gabbouj, M. (2011) · Journal of Signal and Information Processing · doi:10.4236/jsip.2011.22017 · Cited by 8
“Journal of Signal and Information Processing, 2011, 2, 125-139 doi:10.4236/jsip.2011.22017 Published Online May 2011 (http://www.SciRP.org/journal/jsip) Copyright © 2011 SciRes.”
BackgroundMethodInfluentialSubject adaptation using selective style transfer mapping for detection of facial action units
Mohammadian, A., Aghaeinia, H., Towhidkhah, F., & Seyyedsalehi, S. Z. (2016) · Expert Systems with Applications · doi:10.1016/j.eswa.2016.03.023 · Cited by 8
Knowledge Router: Learning Disentangled Representations for Knowledge Graphs
Zhang, S., Rao, S. X., Tay, Y., & Zhang, C. (2021) · Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT) · doi:10.18653/v1/2021.naacl-main.1 · Cited by 8
“A bilinear model is adopted in (Tenenbaum and Freeman, 2000) to separate content from styles.”
MethodScalable Font Reconstruction with Dual Latent Manifolds
Srivatsan, N., Wu, S., Barron, J. T., & Berg-Kirkpatrick, T. (2021) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2021.emnlp-main.244 · Cited by 8
“One of the more philosophically similar approaches to ours is the bilinear factorization model of Freeman and Tenenbaum (Freeman and Tenenbaum, 1997; Tenenbaum and Freeman, 2000) which also learns vector representations for each font and char-”
MethodShohan, S., Zeng, Y., Chen, X., Jin, R., & Shirwaiker, R. A. (2022) · Biosensors and Bioelectronics · doi:10.1016/j.bios.2022.114286 · Cited by 8
Towards more effective distance functions for word image matching
Jain, R., & Jawahar, C. V. (2010) · IAPR International Workshop on Document Analysis Systems (DAS) · doi:10.1145/1815330.1815377 · Cited by 8
“It has been shown in the past that given a collection, style and content can be separated [24] with the help of expectation maximization (EM) algorithm.”
BackgroundA Physics-Based Analysis of Image Appearance Models
Xu, Y., & Roy–Chowdhury, A. K. (2010) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2010.216 · Cited by 8
“We prove that under suitable conditions, the image appearance space can be closely approximated to be multilinear, with the illumination and texture subspaces being trilinearly combined with the direct sum of the motion and deformation subspaces.”
BackgroundCompact Bilinear Deep Features For Environmental Sound Recognition
Demir, F., Şengür, A., Lü, H., et al. (2018) · International Artificial Intelligence and Data Processing Symposium (IDAP) · doi:10.1109/idap.2018.8620779 · Cited by 8
“Bilinear pooling or second-order pooling was first introduced to the computer vision community in [24].”
BackgroundNovel deep neural network based pattern field classification architectures
Huang, K., Zhang, S., Zhang, R., & Hussain, A. (2020) · Neural Networks · doi:10.1016/j.neunet.2020.03.011 · Cited by 8
Mid-level Fusion for End-to-End Temporal Activity Detection in Untrimmed Video
Rahman, M. A., & Laganière, R. (2020) · British Machine Vision Conference (BMVC) · doi:10.5244/c.34.66 · Cited by 8
“To this end, we are inspired by the latest advancement in Visual Question Answering (VQA) that makes efficient use of bilinear pooling [32] based methods to allow for high-level interactions between the different modalities.”
MethodFrom Local to Global: Efficient Dual Attention Mechanism for Single Image Super-Resolution
Zhang, P., & Lam, E. Y. M. (2021) · IEEE Access · doi:10.1109/access.2021.3105726 · Cited by 8
“The bilinear model [36] considers the pair-wise inter-play among features and achieves a richer representation.”
Manifold Estimation in View-Based Feature Space for Face Synthesis across Poses
Huang, X., Gao, J., Cheung, S.-C. S., et al. (2010) · Lecture notes in computer science · doi:10.1007/978-3-642-12307-8_4 · Cited by 8
“In [ 16 ], Tenenbaum et al. first propose the bilinear model to separate Manifold Estimation in View-Based Feature Space 39”
MethodHah, J., Lee, W.-J., Lee, J., & Park, S. (2018) · Computational Intelligence and Neuroscience · doi:10.1155/2018/6465949 · Cited by 8
Modeling the product manifold of posture and motion
Datta, A., Sheikh, Y. A., & Takeo Kanade, A. (2009) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw.2009.5457588 · Cited by 8
“In addition, we compare the proposed approach against global models of the motion manifold obtained using the bilinear model [18] and PCA.”
BackgroundMethodResultInfluentialModeling Inter- and Intra-Patient Anatomical Variation Using a Bilinear Model
Jeong, Y., & Radke, R. J. (2006) · IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw.2006.127 · Cited by 8
“Tenenbaum and Freeman [13] applied bilinear models to several tasks that have two natural independent variations, such as images of the same set of faces seen in different poses or lighting conditions, alphabetical letters in different fonts, and English vowels spoken by different speakers.”
BackgroundMethodHoogendoorn, C., Sukno, F. M., Ordás, S., et al. (2007) · IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2007.4409133 · Cited by 8
“What we wish to do amounts to extrapolation, which is only one of several applications of bilinear models, as is shown in [25].”
BackgroundMethodInfluentialSavvides, M., Heo, J., & Park, S. W. (2007) · Springer eBooks · doi:10.1007/978-0-387-71041-9_3 · Cited by 8
Park, S. W., & Savvides, M. (2006) · IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw.2006.73 · Cited by 8
“Traditionally, to solve the problem of tensor factorization, Tenenbaum and Freeman assume that there are limited numbers of Gaussian mixtures in the distribution of face images, and apply EM algorithm to get parameters of the Gaussian mixture models [2].”
BackgroundMethod
Metadata
- Authors
- Tenenbaum, J. B., & Freeman, W. T.
- Affiliations
- Massachusetts Institute of Technology (US)
- Mitsubishi Electric Research Laboratories (United States) (US)
- Venue
- Neural Computation
- Publisher
- MIT Press
- Published
- 2000-06-01
- Volume
- 12
- Issue
- 6
- Pages
- 1247-1283
- ISSN
- 0899-7667 (ISSN-L)
- 1530-888X
- OpenAlex
- W2170653751
- Semantic Scholar
- 7e85f7d59e37972ec52cbabfef0512588d87f125
- Cited by
- 840
- References
- 61
- Citations by year
- 2026: 2 · 2025: 17 · 2024: 20 · 2023: 31 · 2022: 38 · 2021: 61 · 2020: 58 · 2019: 70 · 2018: 49 · 2017: 39 · 2016: 56 · 2015: 34 · 2014: 27 · 2013: 39 · 2012: 30
- FWCI
- 13.29
- Type
- journal article
- Language
- en
- Open access
- closed
- Locations
- Neural Computation (publishedVersion)
- PubMed (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Neural Networks and Applications
- Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Face and Expression Recognition
- Physical Sciences › Computer Science › Artificial Intelligence › Speech Recognition and Synthesis
- Keywords
- bilinear model, factor models, perceptual learning, speaker accent, singular value decomposition, expectation-maximization algorithm, factor interactions
- MeSH terms
- Algorithms, Face, Humans, Learning, Pattern Recognition, Visual, Speech Perception, Linear Models, Neural Networks, Computer
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
